Ionic Liquids for Carbon Dioxide Capture: Process Selection

Departamento de Ingeniería Química y Quíáímica Inorgánica Universidad de Cantabria (SPAIN)

IONIC LIQUIDS FOR CARBON DIOXIDE CAPTURE: PROCESS SELECTION

ALBO Jonathan, CRISTOBAL Jorge and IRABIEN Ángel

ICEPE 2011 – 2nd International Conference on Energy Process Engineering. Efficient Carbon Capture for Coal Power Plants June 20 – 22, 2011, Frankfurt/Main – Germany 0. Research group 010.1. L ocati on:

CANTABRIA SANTANDER

SPAIN

0.2. SOSPROCAN DEPRO Group:

Development of Chemical Processes and Pollution Control Group

UNIVERSITY OF CANTABRIA E.T.S de Ingenieros Industriales y de Telecomunicación Dpto Ingeniería Química y Química Inorgánica Tel: 942 20 15 90 Fax: 942 20 15 91 http://departamentos.unican.es/quimica/

1/21 1. Introduction

Motivation

Carbon dioxide (CO2) is one of the major contributors to climate change

Electricity generation: 41% of world total CO2 emissions

Europe – 1365.94 mmTCO Europe: 30 % of the electricity from coal 2 World - 12064.64 mmTCO2

Separation/Concentration of CO2 in fossil fuel combustion is required to mitigate climate change 2/21 1. Introduction

1.3. Recovery of CO2 by dispersive absorption (Amine- based scrubbing )

Drops dragging Flue gas Solvent volatilization

Amines

Amines + CO2 (≈90% Eff . )

Total solvent losses: Residual Gas 15 vol.% CO 2 1.5 Kg MEA/1 tonne CO2 captured (38ºC) 3/21 1. Introduction

1.3. Recovery of CO2 by dispersive absorption (Amine- based scrubbing )

Environmental, socia High dependence between streams (G and L) landeconomic Bigger equipment drawbacks Solvent evaporation and drops dragging (loss of solvent and air pollution) Amines as absorption liquid: Toxicity solvent

SlSolven tlosses avoide d inprocessitintensifi ifided

- Drops dragging avoided: membrane contactor

- Solvent evaporation avoided: ionic liquid

ZERO SOLVENT EMISSION PROCESS Luis,P., Garea, A., Irabien, A. J. Membrane Sci ,2009

4/21 1. Introduction

1.4. Process intensification

I. Equipment changes: NON-DISPERSIVE ABSORPTION

Independent control of gas and liquid flow rates Advantages LdkLarger and known gas-liquid i nt erf aci al area Avoid solvent drops dragging

Gas Polypropylene Liquid Bulk of gas Boundary layer membrane boundary layer Bulk of liquid

CO ,g C 2 * C CO2,g

* C CO2,l

CCO2,ll ≈ 0 C´CO2,g

Solvent volatilization

5/21 1. Introduction

II. Solvent changes: LIQUIDS WITH LOW VAPOR PRESSURE (e.g. Ionic Liquids)

Avoid solvent evaporation Advantages Specific solvent for specific applications: e.g. tailor-made ionic

liquids for CO2 recovery

Organic salts with negligible vapour pressure Avoid solvent evaporation

RTIL: Room Temperature Ionic Liquids Cations:

Melting point< Room Temperature 5 2

3 1 1

3 2 Imidazolium (IM) Pyridinium (Py) Structure: cation – anion etc… Anions

Cl, BR, BF4, PF6,FeCl4 CH3SO4, C2H5SO4….

Product design according to the requirements Tailor-made ionic liquids

6/21 1. Introduction

Experimental setup

1. Hollow Fibre Module (Gas-Liquid 1 Contactor, Liquicel, USA)

Membrane material polypropylene [EMIM][EtSO4] Fibre o.d., m 3·10-4 ations itions Composition of CO2 8.7 to 41 cc -4 dd Fibre i .d ., m 22102.2·10 fdfeed gas s tream, N2 Rest to blbalance Fibre length, m 0.115 vol.% Number of fibres 2300 Temperature, K 287± 1 ting con

le specifi -1 2 aa Gas flow rate, LLmin·min 0.01 uu Effective membrane area, m 0180.18 -1 Membrane pore diameter, µm 0.04 Liquid flow rate, L·min 0.05 Mod Oper

7/21 1. Introduction

ItIntensifi ifitication of CO2 non-dispers ive abtibsorption:

The overall mass transfer coefficient, Koverall:

∆ylm ⋅ PT N = K overall ⋅ CO2 ,g R ⋅T

1,5E-07 Diethanolamine (DEA) -1 -7 NCO2(l)·R·T·PT = 3,690·10 . ∆γlm (-) 2 -1 ) 1,3E-07 R = 0, 9932 Koverall (m·s ) -1 ~3·10-7 (m·s 1,0E-07 -1 T [2] Rongwong W., et al. Sep.Purif. ·T·P 7,5E-08 R R Technol., 2009. ·

CO2(g) 5,0E-08 Amines N

2,5E-08

0,0E+00 0,0E+00 4,0E-02 8,0E-02 1,2E-01 1,6E-01 2,0E-01 2,4E-01 2,8E-01 3,2E-01 3,6E-01 Ionic liquid ∆γlm (-)

1-Ethyl-3-methylimidazolium -1 ethylsulfate Gas feed composition Koverall (m·s ) -7 ([EMIM][EtSO4] or EMISE) 8.7 – 41 vol.% CO2 / N2 (3,69 ± 0,18)·10

[1] Albo et al., Ind. Eng. Chem. Res. 49, 11045-11051, 2010. 8/21 1. Introduction

Ionic liquids (ILs) are compounds of high interest for industry because of their attractive properties as solvents.

Experimental CO2 solubility of gases on ILs is required but..

Need to evaluate their properties by estimative methods

Estimation by means of structure- Effect of different combinations of activity method (QSARs) groups in the CO2 solubility on ionic liquids

Group contribution methods have been applied previously to predict ionic liquids properties, among them…

TiiToxicity esti mati on wi ihVibifihth Vibrio fischery usi ng mol ecul ldar descri ptors [3] Luis P. et al., Ecotox. Environ. Safety. 67, 423-429, 2007.

Ionic liquids water solubility in terms of anion/cation hydrophobicity [4] Rankes et al., Int. J. Mol. Sci. 10 (3), 1271-1289, 2009.

Conductivities, Viscosities, Ostwald solubility and partition coefficient [5] Yansheng et al., Progress in Chemiestry. 21 (09), 1772-1781, 2009. 9/21 Aim

Design a Tailor-made ionic liquid for CO2 capture

Specific aims:

Property estimation by a database including experimental values of CO2 solubility reporting in literature

Evaluation of operating conditions influence on the process

Stu dyof cation/ an ion stttructure iflinfluence on solubilit y

Santander, 2010 2. Methodology

2.1. Database

Composed of CO2 solubility experimental values from literature

Number of data: 4.843 Minimum Maximum s cc Solubility (%wt) 0.014 70.242

cteristi Pressure (bars) 0.09 946 aa

Temperature (ºC) 5 300 a char tt Da Cations Anions Ionic liquids

3 x 29 = 1015

5 11/21 3. Results

313.1. Pressure Pressureinfluenceonsolubility influence on solubility

Cation Anion

Nitrate ([NO3]) Dicyanamide ([DCA]) 1-Butyl-3-methylimidazolium ([bmim]) Tetrafluoroborate ([BF4])

Hexafluorophosphate ([PF6])

ct Trifluoromethanesulfonate ([TfO]) ee

Bis(trifluoromethylsulfonyl) ([Tf2N]) 70,0 [bmim] [NO3] [bmim] [BF4]

ion eff 60,0 [bmim i] ][DCA][DCA] [bmim] [TfO] An 50,0 [bmim] [PF6] Favorable interactions

(bar) 40,0

[bmim] [Tf2N]

ure 30,0 fluoroalkyl-CO ss 2 20,0 Pres Increased size of the anion 10,0 0,0 [6] S.N.V.K.Aki et al., J. Phys. Chem. B, 108, 20355- 0,00 0,20 0,40 0,6020365, 0,80 2004. Mole fraction (x)

12/21 3. Results

3.1. Pressure influence on solubility

Cation Anion

([hmim]) ([C5mim]) Bis(trifluoromethylsulfonyl) ([Tf2N]) ([dmim]) ([emim]) ([C6H4F9mim]) ect ff ([C8H4F13mim])

ion ef 45 [hmim] [Tf2N] tt [C5mim] [Tf2N] Less s iignifi cant th an th e ani on 40 [dmim] [Tf2N] effect Ca 35 [emim] [Tf2N]

(bar) 30

25 [C6H4F9mim] [Tf2N] ure Solubility ss [C8H4F13mim] [Tf2N][Tf2N] P 20

Pres 15 10 5 0 [7] P.J. Carvalho et al, J. Phys. Chem. B, 113, 6803- 6812, 2009 . 0,0 0,2 0,4 0,6 0,8 [8] M.J. Muldoon et al., J. Phys. Chem. B, 111, 9001- Mole fraction (x) 9009, 2007. 13/21 3. Results

3.2. Temperature influence on solubility

Tª Solubility Cation selection is less significant than the anion selection

Anion effect Cation effect

This negative influence is more evident at higher mole fractions of CO2

14/21 3. Results

3.3. Quantitative analysis

Database is restricted for a quantitative analysis at the range of possible operating conditions of CCS

Number of data 64 So lubili ty = H⋅ P

Solubility (H) 0.003-0.85 abase tt Pressure (bars) 1-50 CS Da

CC Temperature (C) (ºC) 25

(A1, A2…, C1, C2…)

Linear regression Descriptors: Presence/ absence anion QSAR Analysis (A) and cation (C)

15/21 3. Results

3.4. Influence on Cation/Anion structure

1 0 ,50 CiC at i ons ‐0,5

‐1 he ETT hea TBP thea bhea bmpy bmim hmpy emim hmim omim dmim DMFH THTDP C5mim hemim hheme VBTMA hmmim F13mim 4F9mim ETO)2IM 44 (( P(VTMA) HH C6 C8H 1 0,5 Anions ‐0,50 ‐1 Cl Ac ISB dca TfO FEP LEV PF6 BF4 TFA SUC SCN FOR PRO NO3 TFES TMA Tf2N IAAC bFEP pFEP EtSO4 lactate CH3SO4 MDEGSO4

CATIONS: (p-vinybenzyl) trimethylammonium [p-VBTMA] CATIONS: N,N-dimethylformamidium Vinvbenzyl trimethylammonium [VBTMA] [DMFH]

ANIONS: tris(heptafluoroethyl) ANIONS: Thyocynate [SCN] trifluorophosphate[FEP] bis(trifluoromethylsulfonyl) [Tf2N] Hexafluorophosphate [PF6]

16/21 3. Results 3.5 . Process operating conditions

1.2 bar 370ºC 40ºC

CO2 PRODUCT P Solubility COMPRESSION 130 bar

CO2 solubility at 40ºC and real process pressures

Solvent Pressure Solubility (X) Ratio IL/MEA MEA (22.7 wt%) 0.639

[bmim][PF6] 1.2 0.11 0.17 ns rocess oo [bmim][Tf N] 0190.19 030.3 pp 2 MEA (22.7 wt%) 1.159

conditi [bmim][PF6] 130 0.89 0.77 S real CC

C [bmim][Tf2N] 10.87

[9] Vranchos et al., Ind. Eng. Chem. Res. 45 (14), 5148-5154, 2006. 17/21 4. General conclusions

CO2 recovery: - Selective absorption using organic solvents (e.g. amines) on equipment based on dispersive absorption (e .g . scrubbers) Since a direct contact between gas and liquid phases takes place:

- Drops dragging SOLVENT LOSSES - Solvent volatilization Development of tailor made ionic liquids for CO2 recovery:

Process intensification: -Cation selection result of less significance when designing the IL 1. Substitution of conventional equitipment for amembrane didevice - Inclusion of fluorinated groups in the IL structure increase solubility value ZERO SOLVENT EMISSION 2 PROCESS - ILscould become strong competitor of MEA at high pressures

2. Substitution of the absorption Ongoing research: Magnetic Ionic solvent for ionic liquids Liquids (MILs)

18/21 5. Ongoing research Magnetic Ionic liquids (MILs)

ILs with anions containing metals complexes 1-Ethyl-3-methylimidazolium Tetrachloroferrate MILs acquire different behavior in the presence of a magnetic field. 1 (1) Response of Emim[FeCl4 ] to a small Nd 2 magnet (0.55 T)

(2) Reported change of N2 bubbles trajectory a) in absence b) in the presence of a magnet

Compounds can have high or low solubility in the MIL depending on the magnetic field applied

[10] Jiang Y., et al. China Particuology, 2007.

Paramagnetic behavior of nbmim[FeCl4] Emim[FeCl4] Antiferromagnetic ordering 1.5

0.04 1 Fe ion) 0.02 µΒ/ 0

-0.02 Fe ion) 0.5 -0.04 Magnetization (

µΒ/ -1000 -500 0 500 1000 Field (Oe)

n ( 0 oo

-0.5

-1 Magnetizati T = 2 K -151.5 -85 -42.5 0 42.5 85 Field (kOe) [11] Hayashi, S., et al. IEEE Transactions on Magn., 2006. 19/21 5. Ongoing research

CO2 solubility measurements using task-specific ionic liquids for CO2 capture

Operating conditions Composition of feed stream, (vol.%)

CO2/N2 20/80 Feed gas flow rate, mL·min-1 50

-1 N2 gas flow rate, mL·min 100 Temperature, K 298,15 Pressure, atm 1 DTGA-60H Shimadzu Sample volume ,µL 30 thermobalance

ss Ionic Liquid Solubility (wt %)

[Emim] [FeCl4] 1,036±0,056

y result [[][Bmim] [FeCl4] 0,953±0,011 tt Inclusion in the analysis 3- 3+ [GdCl6 ][ P66614 ] 4,83±0,54 forHigh a robustCO2 solubility database in -2 2+ [MnCl4 ][ P6661 ] 0,719±0,013 MILs

Solubili 2- 2+ [MnCl4 ] [Aliquat ] 0,807±0,009 3- +3 [GdCl6 ][ Aliquat ] 0,763±0,0038

20/21 6. Future research

Stimuli-responsive membrane system BLE AA SWITCH

A + B A + B

Selective transport of component A (without magnetic field) and selective ttfttransport of component B(iB (in th e presence of magne tic fild)field)

21/21 Departamento de Ingeniería Química y Química Inorgánica Universidad de Cantabria (SPAIN)

IONIC LIQUIDS FOR CARBON DIOXIDE CAPTURE: PROCESS SELECTION

Jonathan Albo

16/16 Departamento de Ingeniería Química y Química Inorgánica Universidad de Cantabria (SPAIN)

Grupo Desarrollo de Procesos Químicos y Control de Contaminantes

Departamento de Ingeniería Química y Química Inorgánica ETSIIy T, Av da. l os C ast ros s/ n Santander, Cantabria. SPAIN Tel. +34 942 201597 Fax +34 942 201591

«Scientific and technological progress ca

to promote human development in an equitable society»

n not be an end in itself but a means 2. Experimental section (II)

Fac ilitated tottransport of CO2 thro ugh h a SILM using tktask-speifiecific ioni c liquid s for CO2 capteture

Data acquisition system

Power supply

Pressure transducers

8/16 2. Experimental section (III)

Experimental setup Theory

(1)

Feed Permeate

Membrane

Ia: Inlet feed compartment Ip: Inlet permeate compartment (2) Oa: Outlet feed compartment Op: Outlet permeate compartment

Figure 8. Experimental setup for CO2 transport using SILMs.

PVDF porous membrane Vacuum chamber (1hr) Soaking in the ionic liquid (2hr)

Table 2. Membrane features.

Membrane type Material Pore size (µm) Thickness (µm)

VERICEL (Pall Corporation) Hydrophilic PVDF 0,20 129 9/16 3. Results: Ongoing research

Partial pressure dependence with the solubility for the ILs selected:

1. [Bmim][Ac] 3- 3+ 2. [GdCl6 ][ P66614 ] ILs behavior

3. [Emim] [FeCl 4]

Solubili ty CO Permea bility N Permea bilitbility SlSelec tiittivity Ionic Liquid 2 2 (%) x1011 (m2/s) x1011 (m2/s) (-) [Emim][Ac] 8,6±0,04 42,69±0,22 1,006±0,007 42,43 [Bm im][A c] 7,70±0,06 30,64 ±0,064 0,893 ±0,003 34,31 3- 3+ --- [GdCl6 ][ P66614 ] 4,83±0,54

[Emim] [FeCl4] 1,036±0,056 13,95±0,25 0,923±0,003 15,11

CO2 solubility data in ILs CORRELATION CO2 /N2 permeability in SILMs

12/16 5. Evaluation of new ionic liquids: ongoing research

Magnetic Ionic Liquids (MILs) as tailor-made CO2 absorption solvents

Development of tailor made MILs for CO2 recovery.

12. Characterization Magnetic behavior. Structural characteristics in magnetic fields with different strengths.

21. MIL Selection Viscosity, solubility, ecotoxicity (EC50) and transport rate.

3.3 MILs for CO2 recovery Study of solubility/selectivity parameters of CO2

4.4 Absorption evaluation Gas stream in membrane contactor with MIL., for non-dispersive absorption.

17/19